Toxic and therapeutic effects of amiodarone in the ... · Toxic and Therapeutic Effects ofAmiodarone in the Treatment of Cardiac Arrhythmias H. LEON GREENE, MD, FACC, ELLEN L. GRAHAM,

1114 lACC Vol 2. No 6December 1983 1114-28

Toxic and Therapeutic Effects of Amiodarone in the Treatment ofCardiac Arrhythmias

H. LEON GREENE, MD, FACC, ELLEN L. GRAHAM, RN, JEFFREY A. WERNER, MD, FACC,

GENA K. SEARS, RN, BRIAN W. GROSS, MD, FACC, JAY P. GORHAM, MD,

PETERJ. KUDENCHUK,MD,GENEB. TROBAUGH,MD

Seattle. Washington

Amiodarone was used to treat cardiac arrhythmias thathad been refractory to conventional medical therapy.The first 70 consecutive patients treated with amioda­rone in this study had at least 6 months of follow-up(range 6 to 24, mean 11) and form the basis for thisreport. Sixty-six patients were treated for ventriculararrhythmias and four for supraventricular tachycardias.

Amiodarone therapy consisted of a loading dose of600 mg orally twice a day for 7 days, and 600 mg dailythereafter. Doses were reduced only if side effects oc­curred. Because of frequent side effects, the dose wasreduced from S72 ± 283 mg per day (mean ± standarddeviation) at 4S days to 372 ± 174 mg per day at 6months. With a mean follow-up of 11 months in the S4patients who continued to take amiodarone, only 4 pa­tients had ventricular fibrillation. Three additional pa­tients experienced recurrent sustained ventricular tachy­cardia in long-term follow-up.

All 70 patients had extensive clinical and laboratoryevaluation in follow-up. Side effects were common, oc­curring in 93% of patients. Thirteen patients (19%) hadto discontinue the medication because of severe side ef-

Amiodarone is an investigational antiarrhythmic agent thathas been recently introduced in the United States (1-4). Itis a benzofuran derivative with weak alpha- and beta-ad­renergic blocking effects and was originally tested as anantianginal drug. It was soon discovered that it had potentantiarrhythmic effects (5-10). Electrophysiologic propertiesof this agent consist of minor slowing of conduction. in­crease in action potential duration and refractoriness of atrial.ventricular and His-Purkinje tissue.

Amiodarone is considered to have little negative inotropiceffect, and the occasional hypotension noted with the drug

From the Center for Arrhythmia and Prehospital Care. Division ofCardiology, Harborview Medical Center. University of Washington. Se­attle. Washington. Manuscript received March 8. 1983; revised manuscriptreceived June 21, 1983. accepted June 22, 1983

Address for repnnts: H. Leon Greene. MD, Center for Arrhythmia andPrehospital Care, Harborview Medical Center. ZA-35. 325 Ninth Avenue,Seattle. Washington 98104.

© 1983 by the Amencan College of Cardiology

feets. Fifty-six patients had gastrointestinal side effects,most commonly constipation. All patients but 1 even­tually developed corneal microdeposits, and 43 patientswere symptomatic. Cardiovascular side effects were un­common. Symptomatic pulmonary side effects occurredin seven patients, with unequivocal pulmonary toxicityoccurring in five. Neurologic side effects, most com­monly tremor and ataxia, occurred in S2 patients. Thy­roid dysfunction occurred in 3 patients, and 32 patientshad cutaneous abnormalities. Miscellaneous other sideeffects occurred in 32 patients.

Amiodarone appears to be useful in the managementof refractory arrhythmias. Because virtually all patientsdevelop side effects when given a maintenance daily doseof 600 mg, lower maintenance doses should be used. Itis unknown if the more severe side effects are dose­related. Amiodarone is difficult to administer becauseof its narrow toxic-therapeutic range and prolongedloading phase. More importantly, the first sign of an­tiarrhythmic failure may be manifest as sudden cardiacdeath.

has been attributed to peripheral vasodilation. One of theprimary advantages of amiodarone is its long half-life. buta disadvantage of this same property is the long time re­quired for the drug to exert its maximal antiarrhythmic effect(5).

Although amiodarone has been tested by many investi­gators and appears to be a good antiarrhythmic agent (II ),its side effects have been reported to be infrequent (12,13).Corneal microdeposits appear in virtually all patients treatedwith the drug for long periods of time (14). Skin rashes areless common. but can include both photosensitive dermatitisand skin discoloration. Thyroid dysfunction (13.15) andpulmonary toxicity (16-19) have been reported. but havebeen uncommon. In addition. minor neurologic side effectshave been noted (20). Side effects seem to correlate withdosage. but the proper dosage for treatment with amiodaroneis unknown.

0735-1097/83/$300

patients had previously undergone coronary artery bypassgraft surgery, 4 had experienced a cerebrovascular accident,and 10 had a history of chronic obstructive pulmonary disease.

All patients referred for refractory ventricular and supra­ventricular arrhythmias were considered for amiodaronetherapy. All patients gave informed written consent to par­ticipate in this evaluation. The protocol was approved bythe Human Subjects Committee at the University of Wash­ington, beginning August II, 1980.

Amiodarone protocol (Table 3). Patients were initiallyscreened with a number of laboratory examinations beforethe institution of amiodarone therapy. Routine examinationincluded history and physical examination, 12 lead electro­cardiogram, chest X-ray examination, 24 hour ambulatoryelectrocardiographic (Holter) monitor, pulmonary functiontests, thyroid function tests, radionuclide ventriculogram,slit lamp corneal examination, routine blood chemistry de­terminations and unnalysis and liver function tests.

Amiodarone therapy started with a dose of 600 mg orallytwice a day administered for 7 days as an inpatient. Themaintenance dose of amiodarone was then 600 mg dailythereafter. Because amiodarone requires such a long loadingphase, patients were commonly left on the conventionalantiarrhythmic drug that had been found to be the mostefficacious in suppressing their arrhythmias to that time,That conventional drug was continued for 4 to 8 weeks afterinstitution of amiodarone therapy, at which time the con­ventional drug was stopped to determine arrhythmia controlwith amiodarone alone.

Amiodarone dosage was decreased on the basis of sideeffects or increased if arrhythmias recurred after the initial7 day loading phase (Fig. 1). Patients exhibiting good ar-

JACC Vol 2, No 6December 19H3.1114--n

We report our experience with both the antiarrhythmiceffects and the common side effects in our first 70 consec­utive patients treated with amiodarone.

MethodsPatient population. Between August 1980 and Decem­

ber 1982, a total of 130 patients were started on amiodaronetherapy because of symptomatic and refractory cardiac ar­rhythmias, All patients had failed to benefit from conven­tional antiarrhythmic agents (quinidine, procainamide, dis­opyramide and propranolol) either because of inefficacy orside effects, Inefficacy of conventional drugs was deter­mined either by spontaneous recurrence of the clinical ar­rhythmia, ventricular tachycardia on Holter ambulatoryelectrocardiographic monitoring or electrophysiologic test­ing by programmed electrical stimulation. Some patientswere not given disopyramide or propranolol because ofpreexisting severe congestive heart failure. Conventionaldrugs had been given in progressively increasing doses untilside effects appeared or until the drugs were found to beineffective at high therapeutic serum levels. Many patientshad also been given other experimental antiarrhythmic agents,and conventional drugs had commonly been used in com­bination with other experimental agents (Table I). Becauseof the reported length of time necessary for amiodarone toexert its full effect. it was commonly reserved as a lastresort drug for use in patients who had failed therapy withother experimental antiarrhythmic agents that have a shorterhalf-life.

In the first 130 patients started on amiodarone, at least6 months of follow-up were available on the first 70 con­secutive patients treated with this drug. These 70 patientsform the basis for this report. Mean follow-up for these 70patients was II months (range 6 to 24),

Patients were referred for a variety of clinical arrhythmias(Table 2). They ranged in age from 32 to 78 years (average59). There were 56 men and 14 women. Fifty of the 70patients had clinical evidence of coronary artery disease; 21

Table 1. Previously Used Ineffective Drugs

Panents

Nu. %

QUinidine 69 99Procamamide 70 100Disopyramidc 32 46Propranolul 20 29Phenytoin 5 7Apnndmc 17 24Mexiletine 13 19Tocarrude 3 4Clofilium 3 4Bretylium 10 14

GREENE ET ALAMIODARONE TOXICITY

Table 2. Clinical Features of 70 Patients

Diagnose,

Cluneal

Coronary artery disease

Cardiumyopathy

Valvular disease

Wolff-Parkinson-Whue syndrome

Mitral valve prolapse

Arrhythmia

Ventncular fibrillation

Single episode

Recurrent

With recurrent sustained

ventncular tachycardiaVentncular tachycardia

Recurrent sustained

Recurrent unsustained

Single episude sustained

Supraventricular tachycardia

1115

Patient,(no )

509632

32257

16

342932

4

1116 GREENE ET AL.AMIODARONE TOXICITY

JACC Vol 2. No 6December 1983 I I 14- 28

Tab le 3. Armodarone Protocol

Months

Before Drug I Wk 2 3 6 9 12 15 18 21 24

History x x x x x x x x x x )( xPhysical examination x x x x x x x x x x x xElectrocardiogram x x x x x x x x x x x xChest X-ray fi lm x x x x xHolter monitor recording x x x x xSlit lampcorneal examination x x xThyroid function tests x x x x x

Liver function tests x xPulmonary function tests x x x x x xRadionuchde ventriculogram x x

AMIODARONE DOSING

1200 ' 0

Figure 1. Amiodarone dosage versus time. Amiodarone was ini­tially started at 1,200 rng orally daily, then reduced to 600 mgdaily . The dose was reduced only if side effects appeared. Dosesare depicted as mean ± standard deviation. Numbers at each valueindicate number of patients.

care unit or in the telemetry unit until their arrhythmiasappeared to be stable.

Spontaneous failures early in the course of therapy (withinthe first month ) were not considered to be indications for achange in therapy. The drug was considered to be a failureonly if arrhythmias recurred when severe subjective sideeffect s limited an increase in dosage or if arrhythmias (suchas sudden cardiac death or sustained ventricular tachycardia)recurred in the presence of moderate to heavy corneal mi­crodeposits of amiodarone which suggested adequate totalbody stores of the drug .

Clinical follow-up. Antiarrhythmic effect was classifiedon the basis of the presence or absence of sudden cardiacdeath, sustained ventricular tachycardia requiring cardio­version . nonsustained ventricular tachycardia causing symp­toms of syncope or presyncope and asymptomatic arrhyth­mias (including nonsustained ventricular tachycardia) asdocumented on Holter ambulatory electrocardiographicmonitoring. Because most patients had rece ived all con­ventional and many experimental antiarrhythmic agents, therewas great pressure to continue amiodarone. Patients whostopped the drug usually had to return to combinations ofpreviously unsuccessful antiarrhythmic agents or proceed tosurgica l therapy. Therefore, in general, any side effects thatcaused discontinuation of the drug were severe. Side effect swere classified as mild if no dosage adjustment of amio­darone was necessary, moderate if the dose had to be low­ered and severe if the drug had to be temporarily or per­manently discontinued.

ElectrophysioIogic st udies. Initially all patients hadelectrophysiologic testing before amiodarone administra­tion . Patients were studied in the postabsorptive state, eithernonsedated or lightly sedated with diazepam. One to fourmultielectrode catheters were inserted in the femoral andantecubital veins and positioned fluoroscopically (averageskin dose of radiation = 1.2 rads) . Lidocaine I% localanesthesia was used, and lidocaine levels were measured atthe time of stimulation. In no case was the lidocaine levelin a therapeutic range. Patie nts were anticoagulated with

'4

10 128

Duration of Therapy Im onthsl

6

42

52

4

53200

100

00

cVl+<~-e

'".§

"~ 600 " ......L-~

~ 500 ...""" <,

1400 ........E 300 r-----~---..

rhythmia control without side effects were left on a 600 mgdaily schedule. In patients exhibiting unequivocal side ef­fects , the dose was decreased to 400 mg , then if necessaryto 200 mg per day . Occasionally patient s temporarily dis­cont inued the drug and restarted it at a lower dose . Patient swho had spontaneous recurrence of severe arrhythmias (ven­tricular tachycardia) in the first month after institution oftherapy commonly had the dose increased transiently to 600mg twice a day for another 7 days to 3 months. until ar­rhythmia control was obtained or side effects appeared.

After initial experience with the drug, during which timewe noted elevations in serum quinidine and procainamidelevels concomitant with the institution of amiodarone (2 1),patients routinely had daily determinations of serum levelsof quinidine. procainamide and digoxin , if they were takingany of these medications, and prothrombin time if they weretaking anticoagulants.

Follow-up examinations were performed freque ntly (Ta­ble 3). Patients were cont inuously monitored in the coronary

l ACC Vol. 2. No 6December 1983 \ 1\4- 28

GREENE ET AL.AMIODARONE TOXICITY

1117

heparin. Right ventricular pacing was performed in at leasttwo sites and at least three basic cycle lengths, usually 600,500 and 450 ms. Single and double ventricular extrastimuliand burst pacing were used in an attempt to induce ven­tricular tachycardia.

The first patients in the series were tested only a fewdays or weeks after institution of amiodarone therapy beforeit was realized that a longer loading phase might be nec­essary for full drug effect. Subsequent patients were testedonly after documentation of the presence of corneal micro­deposits. usually 2 to 3 months after starting the therapy.One patient was studied less than I week after institutionof therapy; 17 patients were studied I or more months. 13patients 2 or more months and 8 patients 3 or more monthsafter institution of therapy.

Ambulatory electrocardiographic monitoring. Holtermonitoring was performed with two channel Oxford Med­ilog recorders and analyzed with a Reynolds Pathfinder IIsystem. The system was interfaced with a Honeywell 1856­A fiberoptic recorder for continuous printout of all electro­cardiographic complexes. All episodesof ventricular tachy­cardia were played in real time for further analysis. Thetechnician has been documented to be 95% accurate usingthis system as compared with hand-counted tapes played inreal time.

Arrhythmias were classified by severity. f requency andpercentage of half-hour intervals containing each arrhyth­mia . Ventricular tachycardia was defined as three or moreconsecutive ventricular complexes at a rate of greater than100 beats/min.

Electrocardiograms were evaluated in detail fo r the fo l­lowing variables: sinus rate, PR and QT intervals, QRSduration, presence of U waves and arrhythmia.

Radionuclide ventriculograms. Standard techniqueswere used for analysisof radionuclide ventriculograms. Theradionuclide used was technetium-99m pertechnetate withstannous pyrophosphate-labeled red blood cells, 25 mCi perdose, yielding a dose of 0.42 rad (total body) and 1.38 rad(blood). These tests were analyzed for both right and leftventricular ejection fractions. Reproducibility of this tech­nique in our laboratory of two left ventricular ejection frac­tions acquired and measured on the same day is ± 0.03(standard deviation).

Pulmonary function tests. Pulmonary function tests wereperformed by standard techniques. Measurements routinelyobtained were total lung capacity. diffusioncapacity, forcedvital capacity (FVC), residual volume, I second forced ex­piratory volume (FEVd and FEV\/FVC. Reproducibility inthe pulmonary function laboratory with these tests is ± 5%on the same patient on the same day and ± 15% on thesame patient on different days. All tests were performed inthe same laboratory to assure comparability of measurements.

Thyroid function tests. The following measurementsof thyroid function were performed by standard methods:triiodothyronine (T3) by resin uptake, thyroxine (T4) by

radioimmunoassay and thyroid stimulating hormone (TSH).Slit lamp corneal examinations. Slit lamp corneal ex­

aminations were performed to assess amiodarone cornealmicrodeposition. Corneal microdeposits were classified asabsent, faint, moderate or heavy. Examinations were per­formed at least at 2 months and every 6 months and com­monly at more frequent intervals as clinically indicated.

Hepatic function tests. Standard serum tests for hepaticfunction included total bilirubin, glutamic oxaloacetic trans­aminase (SGOT), alkaline phosphatase. glutamic pyruvictransaminase (SGPT) and lactic dehydrogenase (LDH).

Antiarrhythmic drug levels. Amiodarone levels wereunavailable at the time of the study. After it was recognizedthat there might be an interaction between amiodarone andother drugs, serum trough levels of the quinidine. procain­amide and digoxin were measured every day for the first 3days of amiodarone therapy and every other day thereafteras indicated. Serum levels of these drugs were analyzed bythe standard radioimmunoassay methods. Likewise. pro­thrombin times were measured daily for all patients takingcoumadin until it appeared that the prothrombin time wasstable.

Statistical analysis. Data were compared using Stu­dent' s t test for paired and unpaired data, chi-square analysisand Kaplan-Meier product limit estimates of survival asappropriate. Probability (pl values less than 0.05 were con­sidered significant.

ResultsClinical antiarrhythmic efficacy. In a follow-up of 6

to 24 months (mean II ), recurrences of the sudden cardiacdeath syndrome or symptomatic sustained ventriculartachy­cardia requiring cardioversion were uncommon (Table 4).

Ventricular fibrillation . Three patients died of ventric­ular fi brillation, and another patient had ventricular fibril­lation and was resuscitated successfully. One of these pa­tients (Case 50) had been taking400 mg of amiodarone dailyfor 2 months after the initial loading schedule. She remainedasymptomatic, and slit lamp corneal examination at 2 weeksshowed no deposits. Her ambulatory electrocardiogram con­tinued to show multiform and paired ventricular prematurebeats but no ventricular tachycardia before the ventricularfi brillation. A second patient (Case 6) died in the hospitalof ventricular fibrillation. She received the standard loadingdose of 1,200 mg daily for I week and was then maintainedon 600 mg daily. After 4 weeks, the dose was decreasedto 400 mg daily because of severe constipation. Her ep­isodes of ventricular fibrillation then recurred, and she died3 months after starting the therapy. A third patient (Case40) died suddenly after 9 months of taking 400 mg daily.Her ambulatory electrocardiogram continued to show non­sustained ventricular tachycardia. Higher doses of amio­darone were not tolerated because of constipation and skinrash. Anotherpatient (Case 16) who had ventricularcouplets

1118 GREENE ET ALAMIODARONE TOXICITY

lACC Vol 2. No 6December 1983 1114--28

200 8 NoneSurgery 2 NoneSurgery 4 NoneSurgery 4 None

800* 4 None600t 15 None

Surgery 8 None

400 5 None

400 13 None

Table 4. Arrhythmia Recurrence on Amiodarone

At Time of Arrhythmia

Type of DoselDay Duration ofCase Arrhythmia (mg) Therapy (mo)

6 VF 400 340 VF 400 950 VF 400 264 VF, resus 1200 5 day,16 VT, sust--+VF, resus 800 1418 VT, sust 600 1422 VT, sust 600 1.356 VT, sust 400 55 VT, nonsust- 600 9

symptomatic46 VT, nonsust- 900 3

symptomatic4 VT, nonsust- 400 19

asymptomatic20 VT, nonsust- 600 1.5

asymptomatic

DoselDay(mg)

Subsequent Course

Duration ofTherapy (rno) Arrhythmias

*Quinidine added to regimen. t Propranolol added to regimen.Nonsust == nonsustained: resus = resuscitated; sust = sustained; VF = ventncular fibrillation; VT = ventricular tachycardia; - = patient died

on ambulatory monitoring developed sustained ventriculartachycardia, followed by ventricular fibrillation, from whichhe was resuscitated, An additional patient (Case 64) hadventricular fibrillation on the fifth day of amiodarone load­ing. He was successfully resuscitated, and amiodarone load­ing was continued. He was monitored for 3 additional weeksbefore hospital discharge, and has been well without re­currence of arrhythmias since then.

Sustained ventricular tachycardia. Three other patientshave had a recurrence of sustained ventricular tachycardia,The arrhythmias were slow enough to be well toleratedhemodynamically but required electrical cardioversion. Allof these patients had corneal microdeposits, and they hadbeen on the drug for 5, 1.3 and 14 months at the time oftheir recurrence, One patient then had the dose increasedfrom 400 to 800 mg daily and subsequently had quinidinereadded to his regimen without further recurrence of hisarrhythmias. The other two patients, who had been takingamiodarone, 600 and 800 rng/day, could not have the doseincreased further. One of these two patients had posturalhypotension and ataxia with a higher dose, and the otherpatient developed a peripheral neuropathy, proximal muscleweakness and abnormal liver function tests requiring ter­mination of the therapy. Both of these latter two patientssubsequently had surgery for their arrhythmias.

Nonsustained ventricular tachycardia. Four patients hadrecurrence of nonsustained ventricular tachycardia, Two ofthe patients were asymptomatic from this rhythm, but theother two patients had near syncopal episodes with theirarrhythmias. One of the symptomatic patients (Case 5) couldnot have the dose increased further because of constipationand symptomatic corneal microdeposits. Propranolol was

added to this patient's regimen. The other symptomatic pa­tient (Case 46) had neurologic side effects and posturalhypotension at higher doses of amiodarone, so cardiac sur­gery was performed with electrophysiologically-directedaneurysm resection. Of the two patients who were asymp­tomatic from their nonsustained ventricular tachycardia, onepatient could not have the dose of amiodarone further in­creased because of symptomatic heavy corneal microde­posits, and the other patient had to have his dose decreasedbecause of worsening of pulmonary function tests suggest­ing early pulmonary toxicity. None of these four patientshas had recurrence of nonsustained ventricular tachycardia,sustained ventricular tachycardia or sudden cardiac death.

Deaths from nonarrhythmic causes. Nine further pa­tients have died of nonarrhythmic causes during follow-upwhile taking amiodarone. Patient 2 had further progressionof his severe heart failure and died in cardiogenic shock 4months after starting amiodarone therapy. His terminal ad­mission was prompted by his decision not to take his di­uretics, rapid accumulation of 20 pounds (kg) of edemafluid and electrolyte imbalance. His terminal rhythm in car­diogenic shock was a relatively slow ventricular tachycardiaat a rate of 110 beats/min. Patient 3 likewise died of car­diogenic shock after slow progression of a low output syn­drome 3 weeks after starting amiodarone therapy. He diedwith a slow idioventricular rhythm and even terminally didnot have either ventricular tachycardia or ventricular fibril­lation. Patient 7 also died from worsening congestive heartfailure with an idioventricular rhythm as the terminal rhythm2 months after starting amiodarone. Patient 8, who hadsuffered two previous myocardial infarctions without chestdiscomfort, developed worsening congestive heart failure

lACC Vol 2. No 6December 19~3 I114-2~

GREENEET ALAMIODARONE TOXICITY

1119

O-!--..,.-.....,--+---+---.+--!--+--...,....LO

Figure 2. Kaplan-Meier life table analysis of recurrence of sus­tained ventncular tachycardia (VT) or ventricular fibrillation (VF).Patients who continued amiodarone are compared with those pa­tient, who stopped amiodarone because of noncardiac side effects.

60

10034

"-------.l-----'---'--+80

103142

40

20

1--\,11 I''_.

\, P'00089

8 ' _

N54 50

f- 100>'0Q)

c: 80~Vl:Jenis 60u,>:; 40s~

3 1 1 1 1

o 90 180 270 360 450 540 630 720

Days of Followup

-ContInued Armodarone- - - Discontinued Amlodarone

Discontinuation of amiodarone because of toxic ef­fects. Thirteen patients discontinued amiodarone becauseof side effects (Table 5). These patients were subsequentlyfollowed up for a mean of 8 months. and four of thesepatients had either ventricular fibrillation or sustained ven­tricular tachycardia after amiodarone was discontinued. Thesepatients could be considered to be of a possible controlgroupfor the amiodarone patients because the medication wasdiscontinued due to noncardiac side effects. not poor ar­rhythmia control. All of these patients were subsequentlystarted on another therapy. 12 patients on the best alternatemedical therapy and I patient on surgical therapy. Thisgroup with severe amiodarone toxicity consisted of 12 pa­tients with ventricular arrhythmias and I patient with supra­ventricular arrhythmias. Of the patients with ventricular ar­rhythmias. there was a 33% incidence of sustained ventriculartachycardia or ventricular fibrillation (4 of 12) in 8.5 months.whereas after I week of loading therapy the amiodaronegroup had only a 13% (7 of 54) incidence of sustainedventricular tachycardia or ventricular fibrillation in II months(p = 0.009. Kaplan-Meier life table. Fig. 2). If only ven­tricular fibrillation after the first week of therapy is consid­ered. 2 (17o/c) of the 12 patients who stopped amiodaronebecause of side effects had ventricular fibrillation in 8.5months follow-up. whereas only 4 (7o/c) of the 54 patientswho continued amiodarone had ventricular fibrillation in IImonths of follow-up (p = 0.2, Kaplan-Meier analysis. Fig.3).

Electrophysiologic studies. In all patients ventriculartachycardia was inducible both before and after amiodarone,with an average cycle length of 269 before and 305 ms afteramiodarone. In an additional two patients, ventricular tachy­cardia rapidly degenerated to ventricular fibrillation. After

and abdominal pain 36 hours before death. The mechanismof his death is unknown, and an autopsy was not performed.

Patient 32 died of probable amiodarone pulmonary tox­icity. though he also had a recent pulmonary embolus andmyocardial infarction. Patient 42 died of progressive un­controllable congestive heart failure. Patient 47 had severehepatic disease before institution of amiodarone therapy.Even before amiodaronetherapy, he had intermittent hepaticencephalopathy. but was also having recurrent ventricularfibrillation. unresponsive to conventional and other exper­imental antiarrhythmic agents. Three months after institu­tion of amiodarone therapy. he died of hepatic encephalo­pathy. His arrhythmias had been controlled. but the dose hadbeen progressively decreased. fearing that it worsened hishepatic function which was marginal even before the drug.Patient 48 died of a pulmonary embolus. Patient 58 died ofprogressive renal failure and electrolyte imbalance. whichwas thought to be unrelated to amiodarone.

Supraventricular tachycardia. Of four patients treatedwith amiodarone for supraventncular arrhythmias. the drugwas discontinued in one (with recurrent atrial fibrillation)due to inefficacy, and in one (with Wolff-Parkinson-Whitesyndromeand atrial fibrillation) because of side effects. Twoother patients (both with Wolff-Parkinson-Whitesyndrome)have continued therapy with good antiarrhythmic control.

• Treatment termmated by patient. physicrans did not attribute thissymptom to armodarone t Finding later determined not to be due toarmodarone.

Sust = sustained: VF = ventncular fibnllanon. VT = ventnculartachycardia.

PeriodWithout

Armodarone Spontaneous

Major Therapy Arrhythmias Since

Case Side Effects (rno) Arrnodarone Discontmuation

I Myoclonus 24 VT. su-t. multiple episode-<} Agranulocytosis" 3 VF

II Herrubalhsrnus 20 None after surgery forVTIYF

12 Fatigue. memory 10" 4 None

13 Wheezing" <} VF

IX Neuropathy. thyroid 5 None after surgery for VT

dysfunction23 Thyroid dysfunction, 10 VT. suvt, multiple epivodes

skin rash25 Pulmonary toxicity 7 None

3X Pulmonary toxrcrty. 3 None

neuropathy.thyroid dysfunction

52 Pulmonary toxicrty 2 None53 Headachev. fangue lJ None

60 Tremor, lJ None

62 Pulmonary toxicity None

Table 5. Patients Who Discontmued Amiodarone TherapyBecause of Side Effects

II 20 GREENE ET AL.AMIODARONE TOXICITY

l ACC Vol 2. No 6December 1983.111 4- 28

O+-..,.-....,.-....,.---r-...,-...,~...,-...,,.Lo

o 90 180 270 360 450 540 630 720

Days 01 Followup

Figure 3. Kaplan-Meier life table analysis of recurrence of ven­tricular fibrillation (VF). Patients who continued amiodarone arecompared with those patients who stopped amiodarone because ofnoncardiac side effects.

before amiodarone administration, the difference was notstatisticallysignificant(0.05 < p < 0.10). Therefore, it wasdifficult to tell which patients would do poorly on amio­darone on the basis of total counts of premature ventricularcomplexes.

Severity of arrhythmia . Likewise, classification of ar­rhythmia was not particularly helpful in determining whichpatients would do well and which would do poorly. Sevenpatients had ventricular tachycardia on the routine follow­up Holter recording. Three of these patients have been freeof ventricular tachycardia. two patients have had either ven­tricular fibrillation or sustained ventricular tachycardia andtwo patientshave had recurrenceof nonsustainedbut asymp­tomatic ventricular tachycardia. Fifty patients have had noventricular tachycardia on their amiodarone routine follow­up Holter recording while taking amiodarone: 45 of thesehave had no recurrence of symptomatic arrhythmias, 4 pa­tients have had either ventricular fibrillation or sustainedventricular tachycardia and one patient has had recurrenceof asymptomatic and nonsustained ventricular tachycardia.Of the six patients with either ventricular fibrillation orsustainedventriculartachycardia duringfollow-up , only twohad ventricular tachycardia on their amiodarone follow-upHolter recording. Two other patients had no arrhythmiaswhatsoever; one patienthad multiform prematureventricularcomplexes and couplets. and one patient had multiformsingle premature ventricular complexes. Of the three pa­tients with nonsustained ventricular tachycardia in followup, two had ventricular tachycardiaon their routine follow­up Holter recordingand one patienthadonly couplets. Threeof the other 48 patients had ventricular tachycardia on theroutine amiodarone follow-up Holter recording and all ofthese 48 patients havebeenfree of arrhythmias, subsequently.

Therefore, from this small group of patients it is difficultto predict from the Holter monitor recording which patientswill have significant arrhythmias on follow-up. There is aweak trend suggesting that patients who have high densitypremature ventricular complexes on the control Holter re­cording and who exhibit a dramatic decrease in the pre­mature ventricular complex counts will do better, whereasthose patients who have fewer premature ventricular com­plexes on control Holter recording and who actually havean increase in their premature ventricular complex countswill do worse.

Electrocardiographic effects. Amiodarone commonlychanged the electrocardiogram in a characteristic manner(Table 6). Heart rate decreased, PR interval lengthened,QRS interval lengthened , QT interval lengthened and largeU waves commonlyappeared. Though patients were usuallytaking other antiarrhythmic drugs at the time of the controlelectrocardiogram that alone would tend to lengthen theelectrocardiographic intervals, these drugs were commonlydiscontinued in the early months of amiodarone therapy.Cessation of these other antiarrhythmic drugs would have

20-Continued Amiodarone- - - Discontinued Amiodarone

80

N

100 5412N

u,>'S 60~

~ 40

i!!tf. 2

evaluation of these initial 18 patients on amiodarone alone,other reports appeared (22,23) suggesting that virtually allpatients could still have ventricular tachycardia induced whiletaking amiodarone. Therefore, because our experience wasconfirmatory of these other reports, we elected to pursueevaluation of these patients more intensively by Holter am­bulatory electrocardiographic monitoring.

Holter ambulatory electrocardiographic monitoring.Because most of these patients were critically ill with on­

going malignantventriculararrhythmias, the monitoring ob­tained before amiodarone therapy was always obtained whilethe patient was being treated with the best antiarrhythmicagent or combination thereof which had been beneficial inthat patient before the use of amiodarone . Commonly thesecontrol recordings were also obtained on intravenous an­tiarrhythmic agents.

Number ofpremature ventricularcomplexes . Eventhoughpatients were taking other antiarrhythmic agents at the timeof the control monitoring , the switch to amiodarone resultedin a trend toward a decrease in arrhythmia. Patients had amean of 1,228 ± 2,677 premature ventricular complexesbefore and a mean of 238 ± 408 after amiodarone therapy(0.05 < P < 0.10). Patients who subsequently had eitherventricular fibrillation or sustained ventricular tachycardiain follow-up on amiodarone had a mean of 296 ± 538premature ventricular complexes on the control Holter re­cording and 347 ± 477 on the follow-up recording (p =

NS). Other patients who had no significant arrhythmias onfollow-up (including those with nonsustained ventriculartachycardia)had 1,377 ± 2,856 prematureventricularcom­plexes on the control Holter recording and 221 ± 404 onthe amiodarone recording (0.05 < P < 0.10). Though thepatients who had no significant arrhythmias on follow-uphad a higher density of premature ventricular complexes

lACC Vol 2, No 6December 1983.1114-28

GREENE ET ALAMIODARONE TOXICITY

1121

tended to shorten the sinus interval, QRS duration and PRand QT intervals.

The decrease in heart rate due to amiodarone was notsymptomatic in any patient, although some patients devel­oped a heart rate of less than 50 beats/min, One patient hadpreexisting profound sinus bradycardia with rates as low as30 beats/min, and the patient's chronic fatigue was treatedwith a permanent pacemaker. However, amiodarone did notworsen her preexisting sinus bradycardia or congestive heartfailure. Another patient had profound sinus bradycardia withjunctional escape rhythms. His junctional rate was com­monly in the 40s, and amiodarone reduced his heart rate tothe low 40s or upper 30s. In addition, his heart rate wasless responsive to exercise than it had been before amio­darone therapy. Two other patients developed mild sinusbradycardia with heart rates unresponsive to exercise.

Effects on ventricular function. Radionuclide left ven­tricular ejection fraction averaged 0.40 ± O. 14 before amio­darone therapy; the minimal value was 0.10 and the max­imal was 0.87. After 6 to 12 weeks of amiodarone therapy,the minimal left ventricular ejection fraction was O. 12, max­

0.86 and the mean 0.41 ± 0.16. This change in ejec­tion fraction was not significant (n = 57, paired t test). Themajority of the other 13 patients not included in this pairedanalysis had amiodarone discontinued or died before it wastime for the follow-up determination of left ventricular ejec­tion fraction.

Right ventricular ejection fraction was less commonlydetermined, being measured before and after amiodarone in30 patients. The mean right ventricular ejection fraction was0.39 ± 0.14 before and 0.39 ± 0.13 after amiodaronetherapy. The maximal and minimal right ventricular ejectionfractions were 0.92 before and 0.15 and 0.64 and 0.16 afteramiodarone, respectively.

Clinical congestive heart failure. Though the mean leftventricular and right ventricular ejection fractions were un­changed, nine patients continued to have clinical deterio­ration of left ventricular function during amiodarone ther­apy. Six of these nine patients had New York Heart As­sociation (NYHA) functional class III congestive heart fail­ure and 3 had class II heart failure before institution of

amiodarone. At some time during the therapy, the dose of

Table 6. Electrocardiographic Effects of Amiodarone

* 2 to 6 months after initiation of amiodarone therapyp = probability.

Heart rate (beats/rnm)PR intervalts)QRS duranonts)QT interval(s)

BeforeAmiodarone

72 ± 120.19 ± 0.030.11 ± 0.0030.43 ± 0.06

AfterAmiodarone*

65 ± II0.21 ± 0.040.12 ± 0.04

0.46 ± 0.07

p

< 0001< 0 001< 0.005< 0001

amiodarone was reduced in all nine patients because of thefear that the drug had contributed to the worsening of leftventricular function. It is uncertain whether amiodarone pre­cipitated worsening of congestive heart failure or if thesepatients were simply manifesting progression of their in­trinsic disease. Changes in ejection fraction in individualpatients could not be explained by the duration of amio­darone therapy or dose of amiodarone.

Thus, congestive heart failure was new or worse in 9 ofthe 65 patients in whom left ventricular ejection fractionwas measured before amiodarone therapy. If the initial leftventricular ejection fraction was 0.35 or less, 7 patients hadnew or worse congestive heart failure and 23 patients re­mained unchanged. If the initial left ventricular ejectionfraction was more than 0.35, 2 patients had worsening ornew congestive heart failure, and 33 patients were un­changed (p < 0.05).

Postural hypotension. Two patients developed somepostural hypotension when amiodarone was combined witheither quinidine or procainamide, and the blood pressurereturned to normal when either quinidine or procainamidewas discontinued.

Therefore, it appears that amiodarone does not consist­ently change left ventricular ejection fraction, though therisk of new or worsening congestive heart failure is greaterin patients whose initial left ventricular ejection fraction is0.35 or less.

Pulmonary complications. Pulmonary function tests inthis entire group of patients were unchanged. Total lungcapacity decreased only from 97 ± 20 to 94 ± 19% ofpredicted; diffusion capacity decreased from 93 ± 25 to 88± 30% of predicted and FEY I/FYC increased from 76 ±10 to 78 ± 9% of predicted. However, none of these changesreached statistical significance. These pulmonary functiontests were performed a mean of 5.6 ± 3.2 months afterinitiation of amiodarone therapy. No other changes weremade in other medications known to affect pulmonary per­formance, and all but one patient had evidence of cornealmicrodeposits by slit lamp examination at the time of thefollow-up measurements. Patients were taking a mean doseof 7.2 ± 2.4 mg/kg body weight.

Syndrome of amiodarone pulmonary toxicity. This syn­drome is poorly defined. It has usually been characterizedas a constellation of cough, dyspnea on exertion, fever,infiltrates on chest X-ray examination and decreased dif­fusion capacity and total lung capacity. It has been describedas a restrictive process, with pathologic features of pul­monary fibrosis. However, it is not entirely a restrictiveprocess, and it can occur acutely.

In our 70 patients, there were five patients with unequi­vocal amiodarone pulmonary toxicity and another six pa­tients who had possible amiodarone pulmonary toxicity (Ta­ble 7). These complications occurred between 3 and 13months of the initiation of therapy without much warning.

1122 GREENE ET ALAMIODARONE TOXI CITY

JACC Vol 2. No 6Dece mber 19M3 1114- 2M

Ta ble 7. Amiodarone Pulmonary Toxicity

Amiodarone Therapy Preamiodarone

Chest X-rayCase Dose (rug/kg) Durat ion (rno) DLCO TLC FIlm*

Patients With Definite Toxrcity

25 6.7 3 77 98 1+32 8.5 13 11 0 68 037 8.5 12 70 96 1+52 9A 8 70 110 1+62 15.6 7 72 93 1+

Patients With Possible Toxicity

20 lOA 9 80 106 1+22 75 6 79 109 1+34 12.0 9 127 107 035 54 6 77 77 0

38 5.9 9 48 78 1+69 5.5 5 92 100 0

* Degree of interstiual changes on baseline chest X-ray film. graded O. I + . 2+ . 3+ and 4 +DLCO = pulmonary diffusion capacity. TLC = tolallung capactty

The syndrome in these five patients was characterized byabrupt appearance of cough and dyspnea on exertion as­sociated with new pulmonary infiltrates and diffuse rales.Patients were often treated either for pulmonary edema orviral or bacterial infection at other hospitals before it wasrecognized that these changes were probably due to theamiodarone. The pulmonary infiltrates were commonly pe­ripheral, involving the lower lung fields. It was difficult todistinguish between pulmonary infection. pulmonaryedemaand amiodarone toxicity, and it was often the subsequentcourse that confirmed the diagnosis: lack of sputum pro­duction, negative cultures, lack of response to antibioticsand low pulmonary capillary wedge pressure.

One patient (Case 32) in this group of 70 pat ients died5 days after presentation with worsening dyspnea. He ex­perienced significant congestive heart failure before this time.and it was thought that many of his symptoms could beattributed to his congestive heart failure. However, diuresisof 2.7 kg of fluid did not improve his dyspnea. and thechest roentgenogram showed progressive worsening duringthis time. Pulmonary capillary wedge pressure was 6 mmHg, and the patient subsequently died when adequate oxy­genation could not be maintained. On postmortem exami­nation he had a huge heart, congested lungs, evidence ofmultiple small pulmonary emboli. a new occlusion of a smallbranch of the right coronary artery. pulmonary changes ofalveolar septal thickening and proliferation of alveolar clearcells filling the air spaces.

Amiodarone pulmonary toxicity occurred either early orlate in patients taking as little as a 200 mg daily maintenancedose. although two patients were taking 600 mg daily. andtwo patients were taking 800 mg daily. However, there weresome clues to predict which patients would develop toxicity.

Pulmonary function tests showed that the baseline diffusioncapacity was lower (80 ± 17%) prior to the drug in thegroup with certain pulmonary toxicity than in the remainingpatients (98 ± 26%). If the baseline diffusion capacity was80% or less of predicted. 47% of these patients had wors­ening of chest X-ray findings. and 25% of patients hadworsening symptoms. For those patients whose baselinediffusion capacity was greater than 80% of predicted. only6% had worsening of chest X-ray findings and II % hadworsening sypmtoms.

In our larger series of all of the fi rst 130 consecutivepatients (which includes the first 70 patients report ed herein detail), there were seven certain cases of pulmonary tox­icity and six possible cases. The two additional cases ofcertain pulmonary toxicity in the second series of 60 patientswere observed in patients who had been taking 7.5 and 8.5mg/kg. respectively. amiodarone maintenance dose daily.They had been treated with the drug for 2 and 3 months.respectively. Initial diffusion capacity was below 80o/e inboth. 79% in one and 70% in the other. One of these twopatients died. He presented with what appeared to be pul­monary edema that did not clear with vigorous diuresis.After diuresis. there were peripheral lower lung field infil­trates on X-ray examination persisting for weeks with dif­fuse rales. Amiodarone therapy was stopped. Recurrentven­tricular tachycardia had returned approximately 6 weeksafter the amiodarone had been stopped, and 2 weeks laterthe patient died of progressive congestive heart failure ashe was being treated with other antiarrhythmic agents. Onpostmortem examination he likewise had alveolar septal andinterstitial thickening compatible with pulmonary fibrosisdue to amiodarone. In addition. he had diffuse proliferationof alveolar cells obliterating many of the air spaces.

JACC Vol 2. No 6December 1983 1114-28

Table 8. Thyroid Function Tests

GREENE ET ALAMIODARONE TOXICITY

1123

After Arruodarone

Before LastArruodarone I Mo p 3 Mo p 9 Mo p Value* p

T. (4.1 to 11.3 J,lg%t ) 7.0 ± 1.7 8.8 ± 2.3 < 0.001 8.4 ± 3.0 < 0.02 8.5 ± 3.0 < 0.02 8.6 ± 2.5 < 0.001TJ (33 to 45%t ) 42 ± 6 40 ± 6 NS 40 ± 6 NS 39 ± 6 < 0.05 40 ± 6 NSTSH « 6.5 J,lUlml*) 7.4 ± 8.7 13.1 ± 17.5 NS 21.5 ± 36.9 < 0.D25 18.8 ± 26.3 < 0.05 13.8 ± 19.0 < 0.025

* Final determination in follow-up. t Normal value.NS = not significant; p = probability value compared with data obtained before armodarone: T4 = thyroxine measured by radioimmunoassay; TJ

= triiodothyronine measured by resin uptake; TSH = thyroid stimulating hormone.

In all patients who had definite or suspected pulmonarytoxicity. amiodarone administration was at least transientlydiscontinued . In one patient, amiodarone was reinstitutedat 100 mg daily 2 months later when his evidence for pul­monary toxicity had disappeared and arrhythmias began torecur. This patient remainson that dose, free of arrhythmiasand free of evidence of recurrence of pulmonary toxicityfor 2 months.

Thyroid function. Table 8 summarizes the effect of long­term amiodarone therapy on thyroid function . Before amio­darone, serum thyroxine (T4 ) was 7.0 ± 1.7 f.Lg%, in­creasing consistentlyin most patients by I month of therapyand remaining elevated. The serum triiodothyronine (TJ ) atcontrol was normal and decreased slightly. signifi cant onlyat 9 months. The thyroid stimulating hormone was slightlyelevated in the group as a whole before arniodarone, and itincreased progressively throughout the duration of therapy.It is common that this hormone is slightly elevated in pa­tients with severe debilitating disease, but it increased dra­matically in some patients, greater than 100 f.LU/ml in threepatients.

Clinicallyevident hypothyroidism was seen in only threepatients. with some symptoms questionably related to hy­pothyroidism in an additional three patients. No patientsdeveloped clinically evident hyperthyroidism.

Corneal microdeposits. Figure 4 illustrates the per­centage of patients whose slit lamp examinations were pos­itive at various times during follow-up. Almost all patientseventually developcorneal microdeposits, with the majorityof patients having deposits visible between the second andthird months . The earliest deposits were seen at 5 weeks,and the only patient who did not develop corneal micro­deposits on long-term follow-up was also the only patientin the series who wore contact lenses. Ocular side effectswere common, but mild. Patients most frequently com­plained of halos visible around bright lights at night. andsome patients also complained of a gritty sensation in theeyes or itching, dryness or photophobia.

Hepatic function tests. Though there was a trend to­ward elevation of serum enzymes, the variability from pa­tient to patient was quite dramatic, and there were no sta-

tistically significant differences between values before orafter amiodarone therapy. Serum bilirubin was unchanged:0.5 ± 0.3 mg% before and 0.6 :±: 0.4 mg% after arnio­darone. Serum alkaline phosphatase rose only from 82 ±36 to 99 ± 57 Uzliter. SGOT rose from 22 ± 13 to 93 ±391 Uzliter, but the standard deviation was large enoughthat this value did not reach statistical significance at the p< 0.05 level. SGPT increased from 26 ± 20 to 36 ± 21U/liter, and LDH increased from 227 ± 80 to 353 ± 342U/liter. Many of the baseline hepatic function values wereelevated. perhaps in some cases because patients wereundergoing repeated DC cardioversion and resuscitation im­mediately before the initiation of amiodarone.

No patient developed clinically significant hepatic dys­fun ction . One patient who started with hepatic precomaexhibited further deterioration of his hepatic function overthe course of therapy. It is uncertain whether this deterio­ration was secondary to the natural course of his disease or

Figure 4. Percent of patients with corneal microdeposits versustime. Most patients had deposits at 3 months.

AMIODARONECORNEAL MICRODEPOSITS

100

9

E 80III~ 70Q.

E 60~~ 50III

~ 40:;:;'iiiIf 30

~ 20l:!~ 10

1 2 3 4 5Months of Therapy

1124 GREENE ET AL.AMIODARONE TOXICITY

lACC Vol. 2. No 6December 1983:1114-28

the amiodarone. His initial bilirubin was 1.6 mg%, increas­ing to 2.9 mg% after 2 months of therapy. Alkaline phos­phatase was 185 U/liter, increasing to 278 U/liter after ther­apy. SaOT increased from 24 to 32 U/liter, and SaPT wasunchanged at24 Ulliter before and22U/liter afteramiodarone.

Drug interactions. Concomitantadministration of othercardiac drugs was common in patients treated with amio­darone. Because many patients were having on-going ar­rhythmias, conventional antiarrhythmicagents were usuallycontinued during the amiodarone loading phase to provideat least an element of antiarrhythmic protection. We foundinteractions with a number of other cardiac drugs. as re­ported elsewhere by us (21)and byotherinvestigators (24-30).We found that amiodarone decreases dosage requirementsfor digitalis. quinidine, procainamide and Coumadin (Table9).

Only one patient whose digitalis levels were elevateddeveloped symptoms and signs of toxicity which includednausea, vomiting and bradycardia. Though patients com­monly demonstrated electrocardiographic manifestations ofelevated quinidine or procainamide levels (widened QRScomplexes. long QT intervals or bradycardia) , only fivepatients developed symptoms of quinidine toxicity and onedeveloped symptomsof procainamidetoxicity. No bleedingoccurred with prolongation of prothrombin time, but dos­ages of Coumadin were adjusted promptly.

Subjective side effects. Table 10outlines the subjectiveside effects experiencedby patients, classifiedas mild, mod­erate or severe. Most common subjective side effects wererelated to the gastrointestinal tract. Constipation or obsti­pation occurred in 39 patients. Seven patients had nausea,five had anorexia and five had ageusia.

Neurologic symptoms most commonly consisted of tremor(17 patients), ataxia (11 patients) and dizziness (8 patients).At times these symptoms were devastating, requiring ter­mination of the drug. A few patients experienced a varietyof other neurologic side effects including memoryloss, ver­tigo, headache, myoclonus, hemiballismus and peripheralneuropathy.

Cutaneousmanifestations wereseen in 32 patients, whichusually consisted of photosensitivity or itching. Patients

Table 9. Drug Interactions With Amiodarone

No. % Increase % Dose

Drug Patients in Serum Level" Reduction Required f

Digitalis 4 280 50Quinidine II 32 37Procainamide 12 57 20Coumad in II 43

" At constant drug dose. t To maintain constant serum level.

commonly had to use a sunscreen to prevent sunburn, andsome patients described a " crawling" or "prickling" sen­sation beneath the skin. One patient developed hyperkera­toses, and another two patients were aware of a dark blueskin discoloration. A milder slate blue discoloration wasevident to physicians in two additional patients, but it wasnot noticed by the patients or their families.

Other subjective side effects included epididymitis in sevenpatients. requiring temporary discontinuation of amiodaronein four patients and subsequent dose reduction in all.

Thus , ofthe 70 patients reported in this series, 65 patientshad subjective side effects . Only one patient in the entireseries has both continued to take amiodarone without everdevelopingany subjectiveside effects and has remained freeof arrhythmias.

DiscussionThe ideal antiarrhythmic agent is still unavailable. An

ideal agent should be efficacious, well tolerated with fewside effects, have a long half-life and infrequent dosingintervals, a high therapeutic to toxic ratio, should not affectother organ systemsor be affectedby diseasesof other organsystems and should be relatively inexpensive. Amiodaronefails these strict criteria because of its frequent side effects.

Dosage of amiodarone. Sideeffects of arniodarone, givenaccording to the dosage schedule outlined, are almost uni­versalbut fortunately often minor. At the time that this studywas initiated, recommendations for amiodarone dosageweregenerally larger than recommendations made today (31 ,32),and it is possible that many of the side effects could beavoided by using lower doses of the drug. However, choos­ing the proper dose of this drug is difficult. The lowestpossible dose should be used to attempt to limit side effects,but on the other hand, the dose should be kept high enoughfor optimal arrhythmia control. In patients with severe ven­tricular arrhythmias, the first evidence for an inadequatedose can be the appearance of sudden cardiac death. Evenwith the relatively large maintenance dose used in this study(with its attendant side effects) there were six recurrencesof ventricular fibrillation and five recurrences of sustainedventricular tachycardia in the total group (including thosepatients who stopped the drug).

A major limitation of amiodarone is the time requiredfor the drug to become effective. One study (5) reported thatthe antiarrhythmic effect appeared at approximately 9 days.Patients with recurrent, life-threatening ventricular arrhyth­mias need control more quickly, and it is possible that thefull effect of amiodarone cannot be obtained for 1 to 2monthsafter the drug therapyhas been started. Someauthors(33) suggest that this problem may be obviated by startingwith the intravenous drug. Patients in our series had toremain in the hospital on a cardiac monitor until we were

JACC Vol. 2. No.6December 1983:1114-28

Table 10. Amiodarone Side Effects

GREENEET AL.AMIODARONE TOXICITY

1125

Mild Moderate Severe

+ ::': + ::': + ::':

GastrointestinalConstipation 33 4 1Obstipation 1Nausea 3 IAnorexia 1 2Ageusia 2 2

OcularHalo effect 17Blurred vision 5Eye itching. dryness 10Photophobia 8

CardiovascularPostural hypotension IBradycardia 2 3Worsening CHF 6 2

PulmonaryDyspnea on exertion with

pulmonary infiltrates 7

NeurologicHeadache 1Dizziness I 3 3Ataxia 4 5 2Vertigo I 3Short-term memory loss 2 2 2Myoclonus-abdominalTremor 14HemiballismusParesthesias-peripheral neuropathy

EndocrineHypothyroidism 2 2Parotid swelling

CutaneousPhotosensitivity 16

with rash 7Change in skin color 2HyperkeratosesItching 3 2

MiscellaneousInterrupted sleep pattern to 2Fatigue 3 2Depression 2 IExercise tolerance 3 2Epididymitis 7

CHF = congestive heart failure; Mild noticeable. does not interfere with routine activities; Moderate = bothersome. requiring dose adjustmentand/or supplemental therapy; Severe = may necessitate cessation of amiodarone: + = amiodarone-related: ::': = ? amiodarone-related; - = doubtamiodarone-related .

1126 GREEN E ET ALAMIODARONE TOXICITY

lACC Vol 2. No 6December 1983.1114-28

certain that the drug had accumulated sufficiently to have atherapeutic effect, usually a minimum of 7 to IO days. If apatient developed ventricular tachycardia or ventricular fi­brillation again during this time the drug was simply con­tinued at high dose for longer periods of time .

Both initial and maintenance doses of the drug are dif­ficult to establish . The decision that loading has been suf­ficient and that the dose should be reduced is aided by severalclues. Measurement of amiodarone serum levels is generallyunobtainable in most hospitals . Measurement of reverse T3

has been reported to be useful in determining when the drughas accumulated sufficiently (34). However, we most oftenused suppress ion of ambient ventricular arrhythmias andother electrocardiographic effects such as prolongation ofthe PR and QT intervals, widening of the QRS and devel­opment of sinus bradycardia and large U waves as indexesof drug effect. Documentation of corneal microdeposits isanother clue that the drug has accumulated sufficiently, thoughkeratopathy usually occurs after the antiarrhythmic effecthas been achieved. Initially, the maintenance dose of amio­darone in our patients was a mean of 572 mg per day. Morerecently, the mean maintenance dose for patients who havetaken the drug for at least 3 month s has been 327 mg perday .

Our overall approach to the use of the drug has changeddramatically . With the first 70 patients the doses were re­duced only if side effects were either moderate or severe.More recently, we have been routinely decreasing the dosebefore the appearance of side effects in an attempt to main­tain a daily dose of 200 to 400 mg. Patient s with supra­ventricular arrhythmias or less life-threatening arrhythmiasmay be controlled by giving even lower doses and thusfurther minimizing the possibility of side effects.

Antiarrhythmic efficacy. Amiodarone has been used inthe United States primarily for patients with recurrent andrefractory ventricular tachycardia and ventricular fibrilla­tion. Amiodarone seems to be quite efficacious for suppres­sion of ventricular arrhythmia, both simple and complexforms (1-12) . Though convincing survival studies are stilllacking, amiodarone seems to be useful in preventing ven­tricular tachycardia and may have a favorable effect onmortality in patients with the sudden cardiac death syn­drome . Only 4 of 54 of our patients had ventricular fibril­lation on follow-up, compared with 2 of 12 patients whohad to stop amiodarone because of side effects. Though thiscomparison does not constitute a controlled study, at leastthe patients in both group s were selected for use of the drugby the same criteria, the one group having amiodarone dis­continued simply because of noncardiac side effects and notbecau se of antiarrhythmic inefficacy. Obviously, controlledrandomized studie s are needed to define the precise role ofamiodarone in preventing recurrences of sudden cardiac death.

We did not recognize worsenin g of ventricular arrhyth-

mias with amiodaron e in any of our patients as has beenreported recently (35-39). The four patients who died sud­denly during follow-up had been on amiodarone for longperiods of time. Amiodarone probably did not contribute totheir subsequent fatal arrhythmias.

Electrophysiologic studies and Holter ambulatorymonitoring. Our findings confirm the observations of oth­ers (22,23) that arrhythmias are invariably inducible duringelectrophysiologic testing on amiodarone therapy. Most ofour electrophysiologic studies were performed many week safter the patient s had begun the drug, and usually the patientshad corneal microdepo sits at the time of programmed elec­trical stimulation. Thus , amiodarone does not seem to pre­vent the induction of ventricular arrhythmias by pacing tech­niques, although it is possible that certain characteristics ofthe arrhythmia induced might predict subsequent clinicalsuccess or failure of this agent (40,41 ). The finding thatarrhythmias are inducible during amiodarone therapy is con­sistent with the concept that the myocardial substrate thatcould sustain an arrhythmia is still susceptible to catheterstimulation techniques at a time when the spontaneous ar­rhythmias are well controlled (42) . This finding may meanthat amiodarone in some way affects the trigger events orthe initiating beats responsible for the induction of spon­taneous ventricular arrhythmias.

Likewise. determination of a good response to amioda­rone by Holter ambulatory electrocardiographic monitoringwas difficult . Though counts of ventricular premature com­plexes decreased in patient s who had a good response , andthe count s increased in patients who subsequently had arecurrence of severe arrhythmias, the results were barelystatistically significant. The findings overlapped enough be­tween patients well controlled and those who would sub­sequently develop a recurrence of sustained ventriculartachycardia or ventricular fibrillation that we were uncertainwhen to alter therapy on the basis of results of Holter am­bulatory electrocardiographic monitoring. In addition, thepatients with the least arrhythmia on control monitor re­cordings seemed to have the worst outcome, making anassessment of drug efficacy on the basis of suppre ssion ofarrhythmia even more difficult.

Ventricular function. Many of our patients entered thestudy with severe conge stive heart failure. During the periodof follow-up , nine patients had worsening of the preexistingheart failure . However, overall ejection fraction in thesepatients has not systematically deteriorated with long-termamiodarone therapy. It is unknown whether the changes inclinical status in these nine patient s were the result of amio­darone therapy or simply the progression of their underlyingmyocardial disease. Compared with some other new an­tiarrhythmic agents, amiodarone seems remarkably free ofdepressant effects on myocardial function.

lA CC Vol 2. No 6December IY83 1114- 28

GREENE ET ALAMIODARONE TOXICITY

1127

The high mortality rate in our series of patients was mostcommonly caused by nonarrhythmic deaths . Death as a re­sult of congestive heart failure was common. Other causessuch as pulmonary embolism and recurrence of myocardialinfarction further increased the mortality in this patient group.These findings simply emphasize the fact that many of thesepatients had terminal diseases. even though their arrhyth­mias could be well controlled.

Pulmonary toxicity. Pulmonary toxicity remains the mostserious potential common side effect of this drug. Five pa­tients in our series of the first 70 consecutive patients de­veloped unequivocal changes consistent with pulmonarytoxicity. One of these five patients died with pulmonarytoxicity potentiallycontributingto his death. Anotherpatientin the series of the next 60 patients also died with changessuggestive of pulmonary toxicity. and both patients hadpathologic findings consistent with those reported before.

Predicting which patients will develop pulmonary toxicityhas been difficult . Preexisting chronic lung disease withdepression of diffusion capacity has weakly predicted whichpatients will develop pulmonary toxicity on long-term fol­low-up. Pulmonary toxicity can appear abruptly. either earlyor late dunng the therapy. and it commonly masqueradesas viral or bacterial infection or worsening congestive heartfailure. The relation to duration of therapy. maintenancedose and total dose remains unclear. The process appearsto be a restnctive disease With thickening of pulmonaryparenchyma and obliteration of alveolar spaces by prolif­eration of clear cells.

Effects of thyroid gland, cornea and liver. Thyroidfunction tests were commonly altered with an elevation ofthyroxine (T4 ) . depression of triiodothyronine (Tj ) and el­evation of thyroid stimulating hormone. These findings areconsistent with block by the drug of conversion of T4 to T ,.However. only three patients were clinically hypothyroidrequiring thyroid replacement.

All patients eventually develop ed corneal microdepositsexcept fo r the one patient who wore contact lenses. Themicrodeposits appeared as early as 5 weeks, and most pa­tients had developed at least some corneal deposits by 2 to3 months of therapy. Only rarely were these patients sig­nificantly bothered by ocular symptoms.

Changes in hepatic function paralleled those changesreported in other serie s, although they were statisticallvinsignificant in our serie s . Many of our patients were ex­tremely ill at the time the drug therapy was started. eitherwith severe congestive heart failure or recurrent need forcardioversion or resuscitation. or both, so initial serum en­zyme elevations were common in these patients. One patientmay have developed worsening of his preexisting severehepatic dysfunction.

Drug interactions. As previously reported. amiodaroneinteracts witha numberof other medications used commonly

by these patients (21). Elevation of digitalis. quinidine andprocainamide levels have been documented when amioda­rone was added to an otherwise stable regimen. In addition,amiodarone interacts with Coumadin, prolonging the pro­thrombin time. Initial 50% reduction of the dose of all ofthese agents is now routine for us.

Subjective side effects. Subjective side effects occurredIn 930/c of patients taking 600 mg per day, but only 19%of patients had to discontinue the drug because of these sideeffects. In all of these patients in whom the drug was ul­timately stopped. it was first discontinued and then startedat a lower dose in an attempt to continue the therapy.

Gastrointestinal side effects were most common but usu­ally not severe. Neurologic side effects were more devas­tating. causing dose reduction or discontinuation in manypatients. most of them early in the dosing schedule. Cuta­neous and ocular side effects were also bothersome. butrarely limited the use of the drug.

Therapeutic implications. Amiodarone seems to be auseful drug for refractory arrhythmias. Its general utilitywill be limited by the long time period required before thedrug has its full effect. the uncertainty of the proper dosein an individual patient and the potentially serious side ef­fects of a drug that appears to accumulate in many tissues.The drug must be closely monitored in all patients. Usinga maintenance dose of 600 mg per day. virtually all patientswill have side effects. These side effects can be both both­ersome and dangerous (43.44). with an incidence of pul­monary toxicity between 7 and 160/('. Lower maintenancedoses are necessary to minimize side effects. but it is un­known if the more severe side effects can be completelyeliminated by dose reduction. The toxic to therapeutic ratioof armodarone may be closer to unity than previously ap­preciated. Even though these patients are at risk of suddendeath. the therapeutic strategy must be changed to eliminateside effects. Once the arrhythmia seems to be under goodcontrol as measured by elimination of ventricular prematurecomplexes and spontaneous runs of ventricular tachycardia.the dose should be progressively decreased. The physicianshould monitor other factors such as the electrocardiogram,Holter monitor recordings. triiodothyrorune (T , ) and cornealmicrodeposits to maintain an adequate amiodarone effectwithout allowing breakthrough arrhythmias that could ini­tially present as sudden cardiac death.

Finally. randomized controlled studies are still needed todefine the effect of amiodarone on survival. Our comparisonof patients who continued the drug with those who had todiscontinue the drug because of extracardiac side effectswould suggest a beneficial effect of amiodarone on com­bined suppression of recurrent ventricular tachycardia andfibrillation. However. until better controlled survival studiesare performed. amiodarone must still be considered Simplyan unproven experimental agent with great promise.

1128 GREENE ET AL.AMIODARONE TOXICITY

IACC Vol. 2, No 6December 1983 1114--28

ReferencesI. Heger 11, Prystowsky EN, Jackman WM, et al. Arruodarone. Clinical

efficacy and electrophysiology dunng long-term therapy for recurrentventricular tachycardia or ventncular fibrillation. N Engl J Med1981;305:539-45.

2. Waxman HL, Groh WC, Marchlinski FE, et al. Amiodarone for con­trol of sustained ventricular tachyarrhythnuas: cluneal and electro­physiologic effects in 51 patients. Am J Cardiol 1981;50:1066-74.

3. Nademanee K, Hendrickson JA, Cannom OS, Goldreyer BN, SinghBN. Control of refractory life-threatening ventricular tachyarrhythmiasby armodarone. Am Heart J 1981;101:759-68.

4. Podrid PJ, Lown B Amiodarone therapy in symptomatic, sustainedrefractory atnal and ventricular tachyarrhythmias. Am Heart J1981;101:374-9.

5. Kaski JC, GirottI LA, Messuti H, Rutitzky B, Rosenbaum MB. Long­term management of sustained, recurrent, symptomatic ventriculartachycardia with amiodarone. Circulation 1981;64:273-9.

6. Rosenbaum MB, Chiale PA, Halpern MS, et al. Clinical efficacy ofamiodarone as an anti-arrhythmic agent Am J Cardiol 1976;38:934-44.

7. Ward DE, Camm AJ, Spurrell RAJ. Clinical antiarrhythmic effectsof amiodarone in patients with resistant paroxysmal tachycardias. BrHeart J 1980,44:91-5.

8. Marcus FI, Fontaine GH, Frank R, Grosgogeat Y. Clinical pharma­cology and therapeutic applications of the antiarrhythmic agent, arnio­darone. Am Heart J 1981;101:480-93.

9. Leak 0, Eydt IN. Control of refractory cardiac arrhythmias withamiodarone. Arch Intern Med 1979;139:425-8.

10. Rosenbaum MB, Chiale PA, Ryba 0, Elizari MV. Control oftachy­arrhythmias associated with Wolff-Parkinson-White syndrome byamiodarone hydrochlonde. Am J Cardiol 1974;34:215-23.

II Nademanee K, Hendrickson JA, Peterson B, Cannom 0, Hecht H,Singh B. Arruodarone: possibly an ideal antiarrhythmic agent (abstr).Am J Cardiol 1982;49:981.

12. Rotmensch HH, Belhassen B, Ferguson RK. Amiodarone-benefitsand risks In perspective. Am Heart J 1982;104:1117-9.

13. Harris L, McKenna WJ, Rowland E, Holt OW, Storey GCA, KriklerOM. Side effects of long-term amiodarone therapy. Circulation1983;67:45-51.

14 D'Amico OJ, Kenyon KR. Ruskin IN. Amiodarone keratopathy. Drug­induced hpid storage disease. Arch Ophthalmol 1981;99:257-61.

15. Burger A, Dinichert 0, Nicod P, Jenny M, Lemarchand-Beraud T,Vallotton MB Effect of amiodarone on serum triiodothyronine, re­verse triiodothyronine, thyroxin, and thyrotropin. J Clin Invest1976;58:255-9.

16. Sobol SM, Rakita L. Pneumonitis and pulmonary fibrosis associatedWith amiodarone treatment: a possible complication of a new annar­rhythmic drug. Circulation 1982;65:819-24.

17. Marchhnski FE, Gansler TS, Waxman HL, Josephson ME. Amio­darone pulmonary toxicity. Ann Intern Med 1982;97:839-45.

18. Rotmensch HH, Liron N, Tupilski M, Laniado S. POSSible associationof pneumonitis with amiodarone therapy. Am Heart J 1980;100:412-3.

19. Riley SA, Williams SE, Cooke NJ. Alveolitis after treatment withamiodarone. Br Med J 1982;284.161-2.

20. Lustman F. Monseu G. Arruodarone and neurological side effects.Lancet 1974;1:568.

21. SaaI AK, Werner lA, Gross BW, et al. Interaction of amiodaronewith qumidme and procamamide (abstr). Circulation 1982;66(suppl11):11-224.

22. Hamer AW, Finerman WB, Peter T, Mandel WJ. Dispanty betweenthe clinical and electrophysiologic effects of amiodarone in the treat­ment of recurrent ventricular tachyarrhythmias. Am Heart J1981;102:992-1000.

23. Kim SG, Fisher JD, Matos J. Poor predictive value of ventriculartachycardia induced by programmed stimulation In patients takingamiodarone (abstr). PACE 1982;5:305.

24. Moysey 10, Jaggarao NSV, Grundy EN, Chamberlain DA. Arruo­darone Increasesplasma digoxin concentrations. Br Med J 1981;282:272.

25. Hamer A, Peter T, Mandel WJ, Scheinman MN, Weiss D. The po­tentiation of warfarin anticoagulation by amiodarone. Circulation1982;65:1025-9.

26. Nademanee K, Kannan R, Hendrickson JA. Burnam M, Kay I, SinghB. Amiodarone-digoxin interaction during treatment of resistant car­diac arrhythmias (abstr). Am J Cardiol 1982;49: 1026.

27. Martinowitz U. Rabinovici 1, Goldfarb 0, Mang A, Bank H. Inter­action between warfarin sodium and amiodarone. N Engl J Med1981;304:671-2.

28. Rees A, Dalal 11, Reid PG, Henderson AH. Dangers of amiodaroneand anticoagulant treatment. Br Med J 1981;282:1756-7.

29. Southworth W, Friday Kl , Ruffy R. Possible amiodarone-aprindmeInteraction. Am Heart J 1982;104:323.

30. Tartmi R. Kappenberger L, Strinbrunn W, Meyer UA. Dangerousinteraction between amiodarone and quinidine. Lancet 1982;1:1327-9.

31. Nademanee K, Hendrickson JA, Cannom OS, Singh BN. Refractorylife-threatening ventricular arrhythmias: control by arruodarone pro­phylaxis (abstr). Circulanon 1980;62:(suppl II1)II1-151.

32. Fmerman WB Jr, Peter T, Mandel WJ. Studies on the electrophysi­ologic effects of amiodarone in man (abstr). Circulation 1980;62(supplIII):III-152

33. Morady F, Scheinman MM, Shen E, Shapiro W, Sung RJ, DiCarloL. Intravenous amiodarone in the acute treatment of recurrent symp­tomatIc ventricular tachycardia. Am J Cardiol 1983;51:156-9

34. Nademanee K, Singh BN, Hendrickson JA, Reed AW, Melmed S,Hershman J. Pharmacokinetic significance of serum reverse T3 levelsduring amiodarone treatment: a potential method for monitoring chronicdrug therapy. Circulation 1982;66:202-11.

35. Sclarovsky S, Lewin RF, Kracoff 0, Strasberg B, Arditti A, AgmonJ. Arrnodarone-induced polymorphous ventricular tachycardia. AmHeart J 1983;105:6-12.

36. McComb JM, Logan KR, Khan MM, Geddes JS, Adgey AAJ. Amio­darone-induced ventricular fibrillation. Eur J Cardiel 1980;11:381-5.

37. Keren A, Tzivom 0, Gottleib S, Benhonn 1, Stem S. Atypical ven­tricular tachycardia (torsade de pointes) Induced by amiodarone. Chest1982;81:384-6.

38. Guanggeng C, Huang W, Urthaler F. Ventncular flutter during treat­ment With amiodarone. Am J Cardiol 1983;51:609-10.

39. Westveer DC, Gadowski GA. Gordon S. Timmis Gc. Amiodarone­induced ventricular tachycardia. Ann Intern Med 1982;97:561-2.

40. Naccarelli GV, Fineberg N, Zipes DP, Heger 11, Duncan G, Prys­towski EN. Armodarone: discriminant analysis successfully predictschnical outcome in patients who have ventricular tachycardia inducedby programmed stimulation (abstr). Circulanon 1982;66(suppl 11):11­223.

41. McGovern B. Garan H, Malacoff'Rf', DiMarco JP, Sellers TO, RuskinIN. Predictive accuracy of electrophysiologic testing in the treatmentof ventricular arrhythmias with arruodarone (abstr). Circulation1982;66(suppl 11):11-223.

42. Nademanee K, Singh BN Evaluation of antiarrhythmic efficacy ofamiodarone In VT (letter). Am Heart J 1983;105:167-8.

43. McGovern B. Garan H, Kelly E. Ruskin IN. Life-threatening reactionsduring amiodarone therapy (abstr), CIrculation 1982;66(suppl 11):11­224.

44. Nadamanee K, Singh BN, Hendrickson J, et al. Amiodarone In re­fractory life-threatening ventricular arrhythmias. Ann Intern Med1983;98:577-84.